IEEE TRANSACTIONS ON APPLIED SUPERCONDUCTIVITY, VOL. 26, NO. 4, JUNE 2016 4100705

High-Field Solenoid Development for Axion DarkMatter Search at CAPP/IBS

R. Gupta, M. Anerella, A. Ghosh, W. Sampson, J. Schmalzle, D. Konikowska, Y. K. Semertzidis, and Y. Shin

Abstract—Construction and test results of a 100-mm-bore sole-noid made with the high-temperature superconductor (HTS) pro-ducing a peak field of over 10 T are presented. This is the first steptoward building a very high field solenoid (35–40 T)—one of thekey components of the proposed state-of-the-art facility for darkmatter search at the Center for Axion and Precision Physics Re-search, Institute for Basic Science in Korea. The coils made withHTS are expected to provide a field of ∼25 T, and the coils madewith low-temperature superconductors are expected to provide afield of 10–15 T. REBCO HTS tapes were provided by SuNAMalong with the measurements at 20 K in 3.5- to 4-T applied field.The Brookhaven National Laboratory measured several samplesat 4 K in the applied field of 4–8 T. This paper will present aseries of test results of the conductor, six pancake coils, and thefully assembled solenoid.

Index Terms—High-field magnets, HTS, REBCO, solenoids.

I. INTRODUCTION

THE Institute for Basic Science (IBS) Center for Axion andPrecision Physics Research (CAPP/IBS) is setting up a

major facility in Korea for searching for Axion dark matter witha resonant cavity [1]. Axion dark matter is partially converted toa very weak flickering electric field in the presence of a strongmagnetic field. The sensitivity of the detection increases as thesquare of the magnetic field. In addition to a high magneticfield, a large cavity volume is also important. Since the volumeof the magnet dictates the total cavity volume, the key compo-nent of this state-of-the-art experiment will be a magnet withhigh magnetic field and large aperture: 35–40 T and 100 mm.Achieving such high fields requires the use of High Tempera-ture Superconductors (HTS). To reduce cost, it will be a hybriddesign with the outer coil made with conventional Low Temper-ature Superconductor (LTS). LTS coils, likely to be purchasedfrom a commercial vendor, will provide a field of 10–15 Tand HTS coils a field of ∼25 T. The Superconducting MagnetDivision (SMD) at the Brookhaven National Laboratory (BNL)

Manuscript received October 19, 2015; accepted January 6, 2016. Date ofpublication January 25, 2016; date of current version February 8, 2016. Thiswork was supported in part by IBS Korea (project system code: IBS-R017-D1-2015-a00) and in part by Brookhaven Science Associates, LLC under ContractDE-SC0012704, with the U.S. Department of Energy. This work was carriedout under a research agreement between the Institute for Basic Science (IBS),Korea, and Brookhaven Science Associates, LLC under contract No. IBS 2014-B0063 (IBS-NF-14-09).

R. Gupta, M. Anerella, A. Ghosh, W. Sampson, and J. Schmalzle are with theSuperconducting Magnet Division, Brookhaven National Laboratory, Upton,NY 11973 USA (e-mail: gupta@bnl.gov).

D. Konikowska, Y. K. Semertzidis, and Y. Shin are with the Center for Axionand Precision Physics Research, Institute for Basic Science, Daejeon 305-701,Korea (e-mail: yannis@kaist.ac.kr).

Color versions of one or more of the figures in this paper are available onlineat http://ieeexplore.ieee.org.

Digital Object Identifier 10.1109/TASC.2016.2518240

TABLE IMAJOR PARAMETERS OF THE HTS SOLENOID

has recently designed and constructed a REBCO (a secondgeneration HTS) solenoid with a goal of reaching 25 T at 4 K ina 100 mm bore [2] as a part of an R&D module for a Supercon-ducting Magnetic Energy Storage System (SMES). The simi-larity between the two magnets and the experience of designingand building one forms the basis of this program.

II. MAGNET DESIGN

Major parameters of the magnet assembled from six pancakecoils are given in Table I. The goal of the first phase of CAPP/IBS solenoid program was to demonstrate a peak field of atleast 10 T in a 100 mm aperture HTS solenoid with the similardesign and technology that was used in the SMES coil [2] andin the proposed Muon Collider solenoids [3].

The maximum current achieved in the CAPP/IBS solenoidwas 590 A at 4.2 K. This corresponds to a computed peak fieldof 10.8 T on the surface of the coil (see Fig. 1) and a field of8.6 T at the center of the coil, as computed by OPERA [4].

III. CONDUCTOR

A. Specifications

CAPP/IBS pancake coils were wound with the 12 mm wideREBCO tape from SuNAM [5]. Major specifications of theconductor are given in Table II. The conductor was deliveredfrom three production runs, each having a minimum length of140 m without any internal splice. The critical current at 77 K inself-field was high (600 A). The bonding of the copper plating

1051-8223 © 2016 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.

4100705 IEEE TRANSACTIONS ON APPLIED SUPERCONDUCTIVITY, VOL. 26, NO. 4, JUNE 2016

Fig. 1. Computed field contour on the surface of the conductor in the 100-mm-aperture six pancake HTS solenoid built and tested at 590 A.

TABLE IIMAJOR SPECIFICATIONS OF THE HTS TAPE (SUPPLIED BY SUNAM)

Fig. 2. Sample layout of the BNL test holder. The voltage taps at locations Athrough F are indicated.

to the conductor was good as it did not delaminate during thecoil construction and testing.

B. Conductor Measurements

Measurements of samples of the conductor used in theCAPP/IBS solenoid were made both at SuNAM and at BNL.At BNL, samples of the ends of the coil winding were tested at77 K and at 4.2 K. Since the solenoid performance is likely tobe limited by the critical current of the end coils in field perpen-dicular to the wide face of the tape, the measurements at 4.2 Kwere done with field in that direction.

C. Measurement Setup at BNL

The tapes were mounted on a special holder (see Fig. 2 con-nected to high-current leads capable of carrying 1.5 kA at 4.2 Kand can be positioned in a 60 mm bore solenoid that can provide abackground field of 8 T. Voltage taps are positioned 10 mm apart

TABLE IIICRITICAL CURRENT EXTRACTED FROM THE DIFFERENT SECTIONS

Fig. 3. Critical current at 4.2 K versus field applied perpendicular to the wideface of the tape. The variation between the six samples is quite small.

from A to F. There are voltage taps to monitor the joint resis-tance to the HTS leads and across the section that bends aroundto form the straight section which is in perpendicular field.

D. BNL Measurements

The critical current Ic is being reported for a criterion of1 μV/cm at both 77 K and at 4.2 K. The Ic and n-values are de-rived from fitting a straight line to a log E vs. log I , where theelectric field E is in μV/cm. These data at 77 K for a typicalsample are shown in Table III.

The critical current measurements of six samples are shownin Fig. 3. For each sample, the Ic as a function of the externalfield H is fit to a power law Ic=AH−b, where b=0.61±0.015. The average of the six samples is also given. At 8 Tthe variation in Ic is ±19 A, and the lift factor (Ic(4.2 K, 8 T)/Ic(77 K, self-field)) to 8 T is ∼0.82.

The Ic measurements of short samples taken from the endsof the spools show that all conductors are fairly consistent witheach other. The average Ic at 77 K is 644 A, and in perpendicu-lar field of 8 T, 4.2 K is 526 A with a sigma of 3.6%. From thesemeasurements, the performance of individual coils is expectedto be uniform at 77 K and also at higher fields when operated atlower temperatures.

In addition to the measurements performed at BNL, mea-surements were also made at SuNAM on samples taken fromvarious spools. There the measurements were made at 77 K,self-field, and at 20 K, 3.5 T, and 4 T fields applied perpendi-cular to the wide face of the tape. SuNAM provided data foreight of the nine spools, whereas BNL measured six of the ninespools of the conductor. From the data summarized in Table IV,one obtains the following ratios between the critical current at77 K, self-field, to critical current at 20 K, 4 T, to be 0.67 ±0.03, between 4.2 K, 4 T, and 20 K, 4 T to be 1.84 ± 0.07, andbetween 4.2 K, 8 T and 4.2 K, 4 T to be 0.66 ± 0.01.

GUPTA et al.: HIGH-FIELD SOLENOID DEVELOPMENT FOR AXION DARK MATTER SEARCH AT CAPP/IBS 4100705

TABLE IVCRITICAL CURRENT OF SHORT SAMPLES

TAKEN AT BNL AND AT SUNAM

Fig. 4. Computer-controlled automatic winder used for winding CAPP coils.

Fig. 5. (Left) Single-pancake coil with voltage taps. (Right) Double-pancakeunit made with two single pancakes with a spiral splice joint inside.

IV. CONSTRUCTION

Six single pancake coils were wound with 12 mm wide and100 μm thick HTS tape from SuNAM with the automatic coilwinder (see Fig. 4). Turn-to-turn insulation was 25 μm StainlessSteel (SS) tape, which also helps in quench protection and addsto the support structure [2]. As a part of the Quality Assurance(QA) program, a number of voltage taps were installed (see Fig. 5left) and removed after 77 K testing of the individual pancakecoils. Insulating Mylar sheets of 0.25 mm thickness are installedon either side of the pancakes. A double pancake coil unit isbuilt with a spiral joint at the inner radius of two single pancakes(see Fig. 5 right). Three double pancake units were assembled to-gether with copper discs in between and on either ends to makethe CAPP/IBS solenoid. Fiberglass-epoxy and stainless steelband were wrapped under tension on each double pancake toprovide radial support structure (see Fig. 6). Voltage taps wereplaced on the inner and outer radius of each single pancake.Other instrumentation included temperature sensors and a Hallprobe. Design and construction techniques were similar to thoseused in building the SMES coil [2]. The number of turns (whichvary because of the thickness in conductor) and the location ofvarious pancakes in CAPP/IBS solenoid are given in Table V.

Fig. 6. (Left) Fiberglass epoxy and (right) stainless steel band on each double-pancake unit as a part of the radial support structure.

TABLE VNUMBER OF TURNS AND LOCATION OF VARIOUS PANCAKE COILS

Fig. 7. Measured performance (V –I curve) at 77 K (in liquid nitrogen) ofpancake coils #1 and #2 when powered in single-pancake (individually) anddouble-pancake (together) configurations.

V. TEST RESULTS

All double pancake coil units were tested at ∼77 K in liquidnitrogen. Three leads allow testing of individual single pan-cakes in addition to the double pancake coil test. Fig. 7 (V –Icurve) shows the test result of the first unit. The higher criticalcurrent in a single pancake coil test as compared to that in dou-ble pancake coil is due to the lower field on the coil in the for-mer configuration. Figs. 8 and 9 summarize the performancesof all six pancakes in the three double pancake coil tests.

Fig. 10 shows the performance of individual pancakes from“V –I” curve when all six are assembled and powered in seriesat 77 K. Pancake “A” is at the bottom and “F” at the top (seeTable III). As expected, the performance is determined by thepancakes at the two ends (“A” and “F”) due to the perpendicularcomponent of the field. Fig. 11 is “V-T” curve when the currentis held constant at ∼101 A while the HTS solenoid is allowedto warm slowly from 65 K. This method of reaching the criticalsurface has the lowest inductive voltage noise as the current isheld constant. The critical temperature, defined by 1 μV/cm, isreached at 68.3 K. Fig. 12 plots the critical current as a functionof temperature. This plot is the summary of several test runs to

4100705 IEEE TRANSACTIONS ON APPLIED SUPERCONDUCTIVITY, VOL. 26, NO. 4, JUNE 2016

Fig. 8. Measured performance of pancake coils #1 through #6 at 77 K whenpowered in single-pancake configuration.

Fig. 9. Measured performance of pancake coils #1 through #6 at 77 K whenpowered in double-pancake configuration.

Fig. 10. Measured performance (V –I curve) at 77 K (in liquid nitrogen) ofindividual pancake coils when all six are powered together in series.

obtain the critical surface as defined by 1 μV/cm criterion. Themaximum current the CAPP/IBS solenoid reached was ∼590 Awhich corresponds to ∼10.8 T peak field on the coil. Thisexceeds the requirement of demonstrating 10 T peak field.

VI. DISCUSSION

The six pancakeCAPP/IBSsolenoid,which isdescribed in thispaper, is similar to the twelve pancake solenoid built and testedas a part of the SMES [2] program. Both have the same innerand outer radius, and used the same construction techniques. Themain difference was in the conductor. Whereas the CAPP/IBSsolenoid used ∼100 μm thick 12 mm tape from SuNAM having

Fig. 11. Measurement of critical temperature of CAPP/IBS solenoid at ∼101 Awhen the current is held constant and coil is allowed to warm up slowly. Thecriterion of 1.5 mV or 0.1 μV/cm implies a critical temperature of ∼68.3 K.

Fig. 12. Critical current as a function of critical temperature for the criterionof ∼1.5 mV over or 0.1 μV/cm over the entire CAPP/IBS solenoid.

Fig. 13. Magnetic design of a hybrid ∼40-T solenoid. Two inner coils are madewith HTS, and the outer coil is made with LTS.

a ∼40 μm thick copper, SMES used∼160μm thick 12 mm tapefrom SuperPower [6]. The performance of pancake coils madewith SuNAM conductor was very uniform and the conductordid not have any delamination issue. SuNAM conductor has sig-nificantly higher 77 K, self-field critical current but significantlylower lift factor at 4 K and 8 T. Because of this, the maximumcritical current reached in the CAPP/IBS solenoid at 4 K waslower than that in the similar SMES inner coil test.

VII. MAGNETIC DESIGN OF

A VERY HIGH FIELD SOLENOID

A possible magnetic design of a very high field solenoid(∼40 T) is shown in Fig. 13. The inner two layers are madewith REBCO and the outer with Nb3Sn (LTS). REBCO coilscreate ∼25 T field and LTS ∼15 T. The major challenge in such

GUPTA et al.: HIGH-FIELD SOLENOID DEVELOPMENT FOR AXION DARK MATTER SEARCH AT CAPP/IBS 4100705

a magnet will be dealing with very high stresses. Following thedesign of the SMES [2], the coil is divided in three layers tointercept the stress accumulation.

VIII. CONCLUSION

REBCO has a potential for producing very high fields asneeded for the research at CAPP/IBS. Construction and testresults presented here are encouraging. However, significantR&D is still needed given the much higher field and stresses ina 35–40 T, 100 mm solenoid.

ACKNOWLEDGMENT

We would like to thank Glenn Jochen for winding the coilsand Joseph D’Ambra for support provided in testing the sole-noid. We thank SuNAM for providing the measured data on

the conductor. We would also like to thank the administrativestaff at IBS and at BNL for their continuous support during thecourse of this program.

REFERENCES

[1] Center for Axion and Precision Physics Research (CAPP), Institute forBasic Science (IBS), Natural Science Building (E6-2), KAIST, Daejeon,South Korea. [Online]. Available: https://capp.ibs.re.kr/html/capp_en/

[2] R. Gupta et al., “Design, construction and testing of a large aperture highfield HTS SMES coil,” presented at the 24th Int. Conf. Magnet Technology,Seoul, Korea, 2015, Paper 1OrAB_01.

[3] R. Gupta et al., “High field HTS solenoid for a muon collider—Demonstrations, challenges and strategies,” IEEE Trans. Appl. Supercond.,vol. 24, no. 3, Jun. 2014, Art. ID 4301705.

[4] OPERA Simulation Software, Oxfordshire, U.K. [Online]. Available:operafea.com/

[5] SUNAM Co., Ltd., Anseong, Korea. [Online]. Available: www.i-sunam.com/[6] SuperPower Inc., Schenectady, NY, USA. [Online]. Available: www.

superpower-inc.com/